Robust Face Recognition Using Enhanced Local Binary Pattern

Face recognition is an emerging research area in recognition of the people. A novel feature extraction technique was introduced for robust face recognition. Enhanced Local binary pattern (EnLBP) divided the image into sub regions. For each sub region, the salient features are extracted by obtaining the mean value of each sub region. In LBP, each pixel was replaced by applying LBP into each sub region. In this paper, the mean value of sub region was replaced for the sub region. It reduced the dimension of the image and extracts the salient information on each sub region. The extracted features are compared with similarity measures to recognize the person. EnLBP reduces the operation time and computational complexity of the system. The experimental results were carried out in the standard benchmark database LFW-a. The proposed system achieved a higher recognition rate than other local descriptors.